Halide Anion Discrimination by a Tripodal Hydroxylamine Ligand in Gas and Condensed Phases

Physical Chemistry Chemical Physics

Journal: Physical Chemistry Chemical Physics

Manuscript ID CP-ART-07-2019-003764.R1

Article Type: Paper

Date Submitted by the 10-Aug-2019 Author:

Complete List of Authors: Cheisson, Thibault; University of Pennsylvania, Chemistry Jian, Jiwen; Hangzhou Institute of Advanced Studies, Zhejiang Normal University Su, Jing; Los Alamos National Laboratory, Theoretical Division Eaton, Teresa; Embry-Riddle Aeronautical University, Department of Natural Sciences Gau, Michael; University of Pennsylvania, Department of Chemistry Carroll, Patrick; University of Pennsylvania, Chemistry Batista, Enrique; Los Alamos National Laboratory, Theoretical Division Yang, Ping; Los Alamos National Laboratory, Theoretical Division Gibson, John; Lawrence Berkeley National Laboratory, Chemical Sciences Division Schelter, Eric J; University of Pennsylvania, Department of Chemistry

Page 1 of 11 PhysicalPlease Chemistry do not adjust Chemical margins Physics

ARTICLE

Halide Anion Discrimination by a Tripodal Hydroxylamine Ligand in Gas and Condensed Phases Received 00th January 20xx, †,a †,b,d †,c b,e a Accepted 00th January 20xx Thibault Cheisson, Jiwen Jian, Jing Su, Teresa M. Eaton, Michael R. Gau, Patrick J. Carroll,a Enrique R. Batista,*,c Ping Yang,*,c John K. Gibson,*,b and Eric J. Schelter*,a DOI: 10.1039/x0xx00000x

t Electrospray ionization of solutions containing a tripodal hydroxylamine ligand, H3TriNOx ([((2- BuNOH)C6H4CH2)3N]) − − denoted as L, and a hydrogen halide HX: HCl, HBr and/or HI, yielded gas-phase anion complexes [L(X)] and [L(HX2)] . − − Collision induced dissociation (CID) of mixed-halide complexes, [L(HXaXb)] , indicated highest affinity for I and lowest for Cl−. Structures and energetics computed by density functional theory are in accord with the CID results, and indicate that the gas-phase binding preference is a manifestation of differing stabilities of the HX molecules. A high halide affinity of [L(H)]+ in solution was also demonstrated, though with a highest preference for Cl− and lowest for I−, the opposite observation of, but not in conflict with, what is observed in gas phase. The results suggest a connection between gas- and condensed-phase chemistry and computational approaches, and shed light on the aggregation and anion recognition properties of hydroxylamine receptors.

halide binding and recognition have been long-standing13-15 Introduction subjects of interest in supramolecular chemistry.16-23 Although other strategies have been proposed,22, 24 the hydrogen bond Halide anions are prevalent in essentially all aspects of motif has been ubiquitous in these systems. In that context, chemistry, with their use, transport, speciation, and reactivity functional groups such as (thio)urea, amide, pyrrole, or being critical to processes ranging from those in living imidazole have attracted considerable attention due to their organisms1 to nuclear technologies. In the latter case, uranium donor/acceptor properties and geometrical features.15, 22-23 is enriched through its volatile hexafluoride salt,2 plutonium Given our interest in hydroxylamine ligands (R1R2NOH),25-31 legacy-waste contains high chloride concentrations,3-4 while and recognizing their potential for anion binding by means of iodine-129 is an abundant long-lived (half-life = 1.57  107 y) vicinal H-bond acceptors associated with mildly acidic protons, fission product generated in nuclear reactors.5 As a we initiated studies on the propensity of a tripodal receptor consequence, halides represent a substantial fraction of low-, (H TriNOx) for anion capture (Scheme 1). As hydroxylamine intermediate-, and high-level wastes.6-7 Vitrification has been 3 moieties have not been examined in this context previously, it proposed for long-term immobilization and sequestration of was of interest to interrogate their interactions with halides radionuclides.8-10 However, incorporation of a large under a range of conditions, including condensed and gas concentration of halide anions is detrimental to the quality phases, to determine fundamental thermodynamic trends. and sustainability of the formed glass such that these anions must be separated prior to vitrification.6-7, 11-12 Typical Protonation methodologies encompass precipitation, reduction to the H-bond site Ionic radii  acceptors Protic 6 F 1.33 Å 23.3 volatile elemental gas, or anion exchange. On the other hand, HO H-bond donors OH tBu t N Cl Bu HO 1.81 Å 16.3 a. N P. Roy and Diana T. Vagelos Laboratories, Department of Chemistry, University of tBu Pennsylvania, 231 S 34th St., Philadelphia, PA 19104 (USA) E-mail: N N Br 1.96 Å 15.2 [email protected] b. Chemical Sciences Division Weak Lawrence Berkeley National Laboratory H-bond donors I 2.20 Å 13.4 Berkeley, CA 94720 (USA) E-mail: [email protected] H TriNOx (L) c. Theoretical Division, Los Alamos National Laboratory, Los Alamos, NM 87545 3

(USA), E-mails: [email protected]; [email protected] Scheme 1. Structure and characteristics of H3TriNOx (L) and the halide anions 32 휒 d. Present address: Hangzhou Institute of Advanced Studies, Zhejiang Normal considered in this work. Ionic radii according to Shannon; electronegativities ( ) according to Rahm.33 University, 1108 Gengwen Road, Hangzhou, Zhejiang, China, 311231. e. Present address: Embry-Riddle Aeronautical University, Department of Natural Sciences, 3700 Willow Creek Road, Prescott, AZ 86301 (USA) Gas-phase ion chemistry is a versatile technique for † These authors contributed equally. Electronic Supplementary Information (ESI) available: Gas- and condensed-phase obtaining fundamental insights for relatively simple systems methods and characterization, computational and crystallographic details. See absent perturbations encountered in condensed phases.34-36 A DOI: 10.1039/x0xx00000x

Please do not adjust margins PhysicalPlease Chemistry do not adjust Chemical margins Physics Page 2 of 11

ARTICLE Journal Name particularly functional approach is solution electrospray ionization mass spectrometry (ESI-MS) which, coupled to a quadrupole ion trap mass analyser (QIT/MS), can be used to study ion fragmentation.37 Various types of gas-phase complexes and clusters with generic anion-binding interactions of the type [E−Hδ+…X−] (E = N, C, O; X = F, Cl, Br, I) have previously been studied.38-45 In the present work, condensed phase experiments revealed the intrinsic halide (X−) affinity of protonated + H3TriNOx, [L(H)] , to efficiently yield crystalline compounds with the formulae L(HX). Experimental and density functional – − theory (DFT) studies of gas-phase anionic [L(X)] and [L(HX2)] complexes were performed to elucidate the underlying basis for anion recognition by L. Altogether the combined gas and condensed phase studies of L(HX) complexes reveal hydroxylamine as an interesting motif for selective anion Figure 1. Negative-ion mode ESI mass spectra of solutions of L and equal concentrations in ethanol of two halide acids, HXa and HXb: (a) HCl and HBr; (b) binding. HCl and HI; (c) HBr and HI. The L:HXa:HXb ratios are all 1:5:5. The ligand fragmentation patterns are indicated in the structural inset.

Results and Discussion The ESI results for solutions containing I− and either Cl− or Br− (Figure 1b–c), suggest a higher affinity of L for I− as compared ESI Mass Spectrometry. with both Cl− and Br−, at least under these particular ion During an ESI-MS study of binding affinity of L for actinides and production conditions (vide infra). It should also be rare-earth elements,46 our attention was drawn by abundant emphasized that ESI yields do not necessarily reveal solution − anion complexes with compositions [L(HX2)] . Although ESI-MS affinity. For example, ESI may be more sensitive to larger does not necessarily explicitly reveal solution species, the gas- halides such as iodide due to its less effective solvation. It is phase species may indirectly reflect solution affinities. The nonetheless notable that the overwhelmingly dominant ESI − − − observed [L(HX2)] ions were independently and rationally products contain only iodide — i.e. [L(HI2)] and [(L−177)(HI2)] − − formed from solutions of L and acids HX(aq) with X = Cl, Br, I. — with only minor yields of [L(HClI)] and [L(HBrI)] . A Formation of these di-halide adducts motivated the distinctive result is the appearance of substantial − − preparation of gas-phase species such as [L(HXaXb)] with two [(L−177)(HI2)] , where (L−177) indicates an H3TriNOx that has halides, Xa and Xb. Indeed, collision induced dissociation (CID) lost a fragment having a mass of 177 Da. This same of such mixed halide complexes can reveal preferred fragmentation is observed in CID of protonated [L(H)]+ (Figure elimination pathways, that in turn reflect structures and S1) and corresponds to C–N bond cleavage and elimination of energetics that can be directly assessed by computations. The one of the three H3TriNOx “arms” with concomitant back- utility of ESI-MS and CID for assessing structures and bonding transfer of an H atom, as indicated by the purple line in Figure of halide complexes has been described.47-48 1. Although the origins of the characteristic L−177 species are − For Xa= Cl and Xb = Br (Figure 1a), the dominant observed unknown, it suggests a distinctive interaction of I with L. It is gas-phase complexes from ESI were [L(Cl)]−, [L(Br)]−, re-emphasized that such gas-phase species do not necessarily − − − [L(HClBr)] , [L(HCl2)] , and [L(HBr2)] . This nomenclature is not reflect solution speciation. intended to suggest structural or bonding insights, but rather The ESI results demonstrate the formation of gas-phase − only net compositions. The most abundant complexes in [L(HXaXb)] anions and suggest the preferential association of − − Figure 1a, [L(HClBr)] and [L(HBr2)] , contain one or two Br, heavier halides. In order to interrogate this trend we turned to possibly suggesting a higher affinity of L for Br− versus Cl−. The CID experiments on these ions. − gas-phase species [L(HBr2)] may also be present in solution, Collision Induced Dissociation. either as a monomer, or in oligomers that fragment during ESI. − This possibility was assessed by the condensed phase CID performed on the [L(HXaXb)] anions are presented in experiments discussed below. Although small abundances of Figure 2 and show exclusively one CID fragmentation pathway [L(Cl)]− and [L(Br)]− were apparent, ESI resulted in preferential for each of the studied complexes, as given by reactions (1a)– formation of the complexes with two halide anions. (3a): [L(HClBr)]−  [L(Br)]− + HCl (1a) [L(HClI)]−  [L(I)]− + HCl (2a) [L(HBrI)]−  [L(I)]− + HBr (3a)

2 | J. Name., 2012, 00, 1-3 This journal is © The Royal Society of Chemistry 20xx

Please do not adjust margins Page 3 of 11 PhysicalPlease Chemistry do not adjust Chemical margins Physics

Journal Name ARTICLE

same conclusion as reaction (5), though the incremental difference in PAs is ca. 10 rather than 16 kcal mol−1. As discussed below, the observed fragmentation pathways may also be partly a manifestation of structures that favor a particular proton-halide recombination. In view of the neutral HX energetics, the CID results do not necessarily reveal an intrinsically higher gas-phase affinity of − − − H3TriNOx for I over Br over Cl . The computational results below for reactions (1)–(3) yield an assessment of the energetics, as well as possible influence of anion complex structures on favored fragmentation pathways.

Computations: Structures Optimization. To gain insights into the apparently higher affinity of L for heavier halides, we turned to DFT (B3LYP-D3BJ/6-311G**) − − − − − − Figure 2. CID mass spectra of [L(HXaXb)] anions (X = Cl , Br , I ). (a) [L(HClBr)] / calculations. Different possible conformers of [L(HX2)] , the CID amplitude = 0.30 V; (b) [L(HClI)]− / CID = 0.30 V; (c) [L(HBrI)]− / CID = 0.35 V. − − mixed halide complexes [L(HXaXb)] , and [L(X)] (X = Cl, Br, I) were sampled and optimized (Tables S1–S3). Alternative CID pathways, namely reactions (1b)–(3b), − The lowest energy conformer for all three [L(HX2)] complexes described as the loss of HBr(g) as an alternative to reaction (1a) (isomers 1X in Figure 3A and Table S1) presents a “Janus head” and loss of HI(g) as an alternative to reactions (2a) and (3a), conformation. Namely, a halide atom is coordinated by the were not observed. protonated ammonium and a hydroxylamine moiety on one Lighter CID anion products such as bare Cl−, Br−, and I− would side of the receptor, while the second halide is coordinated by not have been detected due to m/z detection limits. In Figures two hydroxylamine groups on the opposite side of the ligand. 2a and 2b, the m/z of unobserved [L(35Cl)]− is indicated in red. Endothermic CID process is governed by two attributes: (i) lower-energy processes are generally favored; and (ii) kinetic barriers may instead favor higher-energy processes. − Elimination of neutral HXa from [L(HXaXb)] presumably + − proceeds by low-barrier association of H with Xa to produce − [L(Xb)] . This hypothesis is supported by computational results which indicate that observed CID pathways are energetically favored and the kinetic barrier is not a determining factor. In addition to relative stabilities of the parent and CID- generated anion complexes, the overall energetics of the observed CID processes, generic reaction (4), incorporate the

formation energy of produced neutral HXb, which is assessed from reaction (5): − − [L(HXaXb)]  [L(Xb)] + HXa (4) Figure 3. DFT-optimized conformers of [L(HCl2)]: “Janus head” form (A) and H + Xa  HXa (5) “Tripodal” form (B). Most significant H-bonding interactions (d < 3 Å) are The energy (kcal mol−1) for reaction (5) is −103 for HCl, −87 displayed as pink dotted lines; other hydrogen atoms were omitted for clarity. for HBr, and −71 for HI.49-50 Neutral HX formation energies from atomic H and X thus favor CID fragmentation to yield HCl A second set of conformers (isomers 2X in Figure 3B and −1 over HBr over HI, in 16 kcal mol−1 increments. Comparatively, Table S1), higher in energy (+5.4 to 5.6 kcal mol ), shows a C3- symmetric “tripodal” conformation where the three formation energies of HX from molecular H2 and X2 are −22.0, −12.4 and −1.2 kcal mol−1 for HCl, HBr and HI, respectively; the hydroxylamine moieties interact with a first halide. On the same stability trend is obtained though with smaller other side, the second halide is stabilized by weak but incremental energy differences. Although the CID results given abundant CH---X interactions. Such short contacts have been by reactions (1a), (2a), and (3a) would appear to suggest a observed, experimentally and computationally, to contribute 23, higher affinity of L for I− over Br− over Cl−, the observed significantly or exclusively to halide binding and recognition. 52-56 pathways could be partially, or perhaps mostly, a Notably the lower energy conformers 1X and 2X for gas- − manifestation of the higher stability of gas-phase HCl over HBr phase DFT-optimized [L(HX2)] units were reminiscent to the over HI. An alternative conceptualization of reaction (4) is from geometries revealed by solid-state crystallography as 51 − discussed below. the perspective of Cooks’ kinetic method, whereby halide Xa − − For the mixed halide compounds [L(HXaXb)] , the “Janus head” or Xb with the higher proton affinity (PA) preferentially retains the proton. Because the order of PAs is Cl− (333 kcal mol−1) > conformers are also energetically favored over the “tripodal” Br− (323) > I− (314),49 this alternative assessment presents the forms (Table S3). In this case, the site selectivity can be interrogated by DFT, the results are depicted in Table S3 and in

This journal is © The Royal Society of Chemistry 20xx J. Name., 2013, 00, 1-3 | 3

Please do not adjust margins PhysicalPlease Chemistry do not adjust Chemical margins Physics Page 4 of 11

ARTICLE Journal Name

− − − Figure 4 for the specific case of [L(HClBr)] . In the most stable example in the reaction [L(I)] + HXb  [L(HIXb)] , the binding − isomers (1a), the heavier halide anions (Xb) are observed to energy of [L(I)] to HXb increases from 34.6 to 37.9 to 40.6 −1 interact with the ammonium and one hydroxylamine arm, kcal mol as Xb becomes heavier from Cl to Br to I. The while the lighter halide (Xa) forms H-bonds with the remaining increase in binding energy from Cl to Br to I seems to be in two hydroxylamine groups in the other side of the receptor. accord with the dominance of the heavier halide complexes However, swapping the two halides in 1a to yield isomer 1b from ESI (Figure 1), but inference of solution speciation from requires only +0.6–1.3 kcal mol−1, with the lowest swapping these spectra must be qualified, as discussed below. − − − − energy for Cl /Br and the highest for Cl /I (Table S2). The reverse of the above complexation reactions for Xa  Xb corresponds to CID fragmentation of the mixed halides − [L(HXaXb)] . The calculated fragmentation reaction energies in − Table 1 and S5 show that for a given [L(HXaXb)] , loss of the −1 lighter HXa requires less energy by 4.410.6 kcal mol . This is illustrated in Figure 4 for the specific case of L(HClBr)]− where

loss of HCl(g) (reaction 1a) over HBr(g) (reaction 1b) is favored by 4.4 kcal mol−1 in accord with the CID results in Figure 2. Besides reproducing accurately the energetics for the net equations (1a), (2a), and (3a); these results also indicate that kinetic barriers do not play a significant role in controlling the final products. From the structures of isomers 1a and 1b of [L(HClBr)]− (Figure 4), it is evident that loss of HBr, which is computed at higher energy than loss of HCl, can only readily proceed from the lowest energy conformer 1a via direct recombination of the ammonium proton and the proximate bromide anion. The lower-energy elimination of HCl is not as Figure 4. Calculated DFT profiles for the CID process (1), DFT-optimized of the directly accessible from this structure. In summary, the different structures are depicted, the most significant H-bonding interactions are − displayed as pink dotted lines. * Interconversion barriers and pathways were not calculation results indicate that for [L(HXaXb)] loss of HXa is computed. favored thermodynamically but not kinetically for Xa lighter than Xb. In order to assess the stability of the products [L(X)]− in − − −1 reactions (1)–(3), different [L(X)] were also optimized (Figure Table 1. Dissociation energies for [L(HXaXb)] (kcal mol ). Bolded values represent the lowest energy pathway. 4 and Table S3). Again, isomers 11X possessing tripodal geometries are located higher in energy (0.5 to 5.0 kcal mol−1) Halides DFT[a] “Intrinsic” [b] than isomers 10X presenting the Janus head conformation. As − − − − − [L(Xa)] + [L(Xb)] + [L(Xa)] + [L(Xb)] + H + depicted in Figure 4, the conformers 10X are binding the [L(HXaXb)]  HXb HXa H + Xb Xa remaining halide with 2 hydroxylamine arms and some CH---X Xa = Cl, Xb = Br 41.5 37.1 118 130 contacts, while the third NOH group interacts with the, now Xa = Cl, Xb = I 45.2 34.6 104 129 neutral, bridgehead nitrogen atom. Notably, the lowest energy Xa = Br, Xb = I 44.0 37.9 103 115 − conformer for [L(I)] is 13I, found slightly below 10I (1.4 kcal mol−1). In this conformer, L adopts a tripodal conformation [a] DFT-calculated energies for the CID pathways. [b] Hypothetical dissociation equation using tabulated energies for HX.49-50 with an internal, intramolecular, H-bonding network—typical of the free ligand25—while the iodide anion is stabilized by Nature of the Experimental CID. multiple CH---I interactions with the benzylic protons (Table − As indicated, the computed fragmentation energies (Figure 4 S3). The increasing relative stability of 13X versus 10X for [L(X)] − isomers from X = Cl to Br to I seems correlated with the and Table 1) demonstrate that fragmentation of [L(HXaXb)] to − decreasing X---H(O/C) interactions as observed in the [L(Xb)] and HXa, where Xa is the lighter halide, is favored by −1 literature.57-58 4.4–10.6 kcal mol . Notably, these energy differences are −1 With these sets of optimized conformers in hand, the different significantly less than the 16–32 kcal mol differences for CID pathways can now be modelled as discussed below. reaction (5) that favor formation of the lighter HXa from the 49-50 association of H and Xa, as discussed above. Accordingly, Computed CID Energetics. the differences in energetics for the dissociation reactions in According to the generic reaction (4), dissociations and Table 1 do not appear to reflect an inherently greater stability − − − − − − of [L(Xa)] versus [L(Xb)] , as for example [L(I)] versus [L(Cl)] . reverse associations ([L(Xa)] + HXb  [L(HXaXb)] ) were Instead, the computed energetics evidently reflects the trend evaluated for Xa = Xb and Xa ≠ Xb. All the association reactions are exothermic, by −34.6 to −45.2 kcal mol−1 (Table S4) in in increasing stability of the produced HX: HI < HBr < HCl. If − − adjustment is made for the relative H + Xa association energies, accord with the low yields of [L(X)] anions over [L(HX2)] − reaction (5), the derived fragmentation energies in Table 1 for observed by ESI (Figure 1). For a given [L(Xa)] , preferential affinity for heavier halides was consistently observed. For reaction (6) actually suggest that the intrinsic stabilities

4 | J. Name., 2012, 00, 1-3 This journal is © The Royal Society of Chemistry 20xx

Please do not adjust margins Page 5 of 11 PhysicalPlease Chemistry do not adjust Chemical margins Physics

Journal Name ARTICLE

increase in the order [L(I)]− < [L(Br)]− < [L(Cl)]−, which is the solid-state crystallography (vide infra). 1H NMR studies in

opposite of what might be casually inferred from the CID CD2Cl2 demonstrated C3v symmetric species on the NMR results in Figure 2. timescale at 300 K. Variable temperature NMR studies − − [L(HXaXb)]  [LXb] + H + Xa (6) revealed broadening of most resonances caused by the The CID results appear to be a manifestation of the higher crystallization of the salts at lower temperature (Figure S29). 1 stabilities of the produced HXa rather than the intrinsic Particularly interesting was comparison of the H NMR spectra − stabilities of the [LXb] . of L(HX), which demonstrates a gradual increase in the shielding of the hydroxylamine protons from L(HF) to L(HI), in Synthesis and Structural Characterization. agreement with the decreasing electronegativity of the In order to provide a condensed-phase basis for comparison respective halides from F− to I− (Figure S30).59 Taken together, with the gas-phase results presented above, we targeted the the spectroscopic data were in accord with the protonation of synthesis of the series of L(HX) (X = F, Cl, Br, and I) compounds. L at the bridgehead nitrogen atom and association of the

Addition of an excess of the HX(aq) acid to an ethanol solution halide anions through H-bond interactions with the of L followed by precipitation in water afforded the salts L(HX) hydroxylamine moieties. The chloride, bromide, and iodide in moderate to good yields (Scheme 2). salts were recrystallized in ethanol and provided suitable crystals for X-ray diffraction (XRD) studies. L(HCl)EtOH and L(HBr)EtOH were isostructural and crystallized as a 1D-H- bonded coordination polymer (Figure 4A for L(HCl)EtOH and S37 for L(HBr)EtOH). In these structures, the [L(H)]+ cations

lack any C3 symmetry and present a “Janus head” conformation noted in the gas-phase DFT-optimized Scheme 2. Synthesis of L(HX) (X = F, Cl, Br, I).. structures: on one side of [L(H)]+, a hydroxylamine group and the ammonium proton interact with a halide anion; on the The isolated salts L(HX) revealed identical ESI products as the other side, an ethanol molecule and the remaining two results presented in Figure 1 (Figures S6–S7). Compounds hydroxylamine moieties are involved in a H-bonding network 1 13 1 L(HX) were characterized by H and C{ H} NMR, infrared (IR) with the anion. The repetition of this motif generates the spectroscopy, solution electrochemistry, and elemental observed supramolecular polymer (Figure 4A). The solid-state analysis, confirming protonation of L to yield the different structure of L(HI)½EtOH was slightly different and consisted of ammonium salts. Going from L(HCl) to L(HI), electrochemical two independent [L(H)]+ units. The first one presents a similar measurement revealed a shift to lower potential for the “Janus head” arrangement as observed in L(HCl)EtOH and oxidation of the hydroxylamine moieties to their nitroxide L(HBr)EtOH, while the second [L(H)]+ acts as a discrete anion (Figure S26). IR spectra of L(HX) revealed a gradual shift to receptor with the “tripodal” configuration (Figure 4B). In this lower frequencies for the NOH stretches going from X = I to F case, the supramolecular chain was permitted by multiple CH-- (Figure S27). This trend and the overall spectra were well- -I(1) interactions from the benzylic protons of [L(H)]+ (H(1b’), reproduced in the predicted IR spectra of DFT-optimized H(12b’), H(23b’) on Figure 4B). Importantly, this motif of − isomers 1X of [L(HX2)] (Figure S28). This suggests large interaction was only noticed for L(HI) and is reminiscent to the structural similarities between the lowest-energy calculated − geometry of the most stable isomer of [L(I)] (13I) obtained − conformers of [L(HX2)] and solid-state L(HX) as confirmed by during the gas-phase DFT optimizations.

This journal is © The Royal Society of Chemistry 20xx J. Name., 2013, 00, 1-3 | 5

Please do not adjust margins PhysicalPlease Chemistry do not adjust Chemical margins Physics Page 6 of 11

ARTICLE Journal Name

Figure 4. Solid-state structures of different salts; heteroatoms are depicted by their thermal ellipsoids, relevant hydrogen atoms are depicted in black, ethanol molecules in light pink. A) H-bonded polymer of L(HCl)EtOH along the a axis. B) Asymmetric unit of L(HI)½EtOH. C) Asymmetric unit of L(HBr)2(CHCl3).

In the previous solid-state structures, an ethanol molecule was that reveal geometries closely related to the DFT-optimized present and directly participated in the H-bond network, structures. raising the question if its presence influences the crystal packing of L(HX). Compounds L(HCl) and L(HBr), recrystallized Solution Speciation. from benzene resulted in displacement of the ethanol Although adducts L(HX) were characterized by NMR molecule while preserving the 1D-H-bonded polymer, spectroscopy and crystallized as supramolecular polymers, we revealing that ethanol was not requisite for the observed were interested in studying more in depth their solution-state crystal arrangement (Figure 5 and S38). speciation to relate with the ESI results. The existence of a soluble, extended polymeric structure is unlikely. When in solution, compounds L(HX) are expected to be present as monomer or small oligomer prior to their crystallization as an extended structure. Chlorinated solvents such as chloroform and dichloromethane were observed to efficiently solubilize (for extended periods of time) L(HCl), L(HBr), and L(HI). 1H

DOSY NMR studies performed on CD2Cl2 solutions of L and L(HBr) revealed similar diffusion coefficients suggesting similar hydrodynamic radii in solution. More importantly, cooling a

solution of L(HBr) in CHCl3 to −20 °C for a week resulted in the formation of single crystals suitable for X-ray diffraction Figure 5. Solid-state structure of polymeric L(HBr) along the b axis as determined by X-ray crystallography. The proposed structures associated with the main ESI studies (Figure 4C). The corresponding solid-state structure species (positive and negative mode) are depicted in red. + demonstrated a nearly C3-symmetric [L(H)] receptor binding to the bromide anion through the hydroxylamine moieties. Despite multiple attempts, crystals of L(HF) suitable for XRD Two chloroform molecules now supplement the coordination characterization were not obtained. In total, L(HX) are easily sphere of the bromide anion. From these observations, we synthesized and form, in the solid-state, H-bonded polymers propose that chloroform or dichloromethane solutions of

6 | J. Name., 2012, 00, 1-3 This journal is © The Royal Society of Chemistry 20xx

Please do not adjust margins Page 7 of 11 PhysicalPlease Chemistry do not adjust Chemical margins Physics

Journal Name ARTICLE

L(HX) consist of monomeric [L(HX)]·n(solv) (n ~ 2–3) with the titration results clearly demonstrate that L(HX) has a [L(H)]+ in the “tripodal” form. strong affinity for binding a second halide, as was also determined by the gas-phase experimental and theoretical Halide Binding in Solution. results. This affinity contrasts with that of neutral L, suggesting Having a clearer picture of the speciation of L(HX) in solution, that protonation pre-organizes a secondary anion receptor. we were interested in assessing the ability to bind a supplementary halide as suggested by the solid-state Gas Phase Oligomerization. polymeric structures and the gas-phase experiments. Addition Referring to the structure of polymeric L(HBr) shown in Figure n of [N Bu4]X to a solution of the respective L(HX) salts (X = Cl, 5, it is evident that the dominant gas-phase species obtained − + Br, I; L(HF) was not sufficiently soluble) resulted in important during ESI, [L(HBr2)] and [L(H)] , directly corresponds to the modifications in the resulting 1H NMR spectra. In particular, indicated units in the 1D-H-bonded polymer. This NOH moieties and one aromatic proton experienced large correspondence between gas and solution prompted us to deshielding (Figures S31–S36). Job’s plots were in accord with revisit the ESI mass spectra in search of larger oligomeric a 1:1 binding model but the limited applicability of this method species. Indeed, substantial abundance of the dimeric species 60 + − prompted us to evaluate alternative models. The binding [(L)2(H2X)] and [(L)2(H2X3)] were observed (Figures S6 and S7), isotherms resulting from 1H NMR titrations at 298 K were which can be represented as [(LH)(Br)(LH)]+ and fitted to several stoichiometries, but were consistently in best [(X)(LH)(X)(LH)(X)]− in direct correspondence to the solid-state agreement with a 1:1 model.61 This was attributed to the structure in Figure 5. Although these oligomers were produced n formation of di-halide adducts [L(HX2)]N Bu4 (Scheme 3). The by ESI, there is no direct evidence that it is present in the −1 − −1 − binding constants KX increased from 245 M for I to 872 M precursor solution, as discussed above. CID of [(L)2(H2Br3)] for Cl (Scheme 3) as observed in related systems.55 Besides (Figure S9) resulted in elimination of neutral L(HBr) to yield n − repeated attempts, single crystals of [L(HX2)]N Bu4 could not [L(HBr2)] . These gas-phase CID results are fully consistent with be grown but it is proposed that these species adopt a “Janus gas-phase species possessing structures and bonding very − head” conformation and to resemble isomers 1X of [L(HX2)] closely related to the solid-state data in Figure 5. Finally, ESI of n obtained by DFT methods (Table S1). solution of [L(HBr2]N Bu4 were similar to the isolated or in-situ prepared L(HBr) confirming that this complex is an adequate H X − X O condensed-phase model for the gas-phase [L(HBr2] adduct tBu H O N n H t [L(HX2)N Bu4] (Figure S8). O Bu H H H H t N n KX = n Bu H + N Bu4X [L(HX)][N Bu4X] N O N H t Condensed Phase Crystallization Process. N Bu n -1 N – N Bu4X tBu N KCl = 872 M O -1 KBr = 479 M In summary, the combination of condensed, gas-phase and in- N H -1 tBu KI = 245 M NnBu O silico methods allows to draw a clear picture of the speciation 4 H X of the tripodal receptor L when contacted with halides or HX X n H O N Bu4 acids. Neutral L has very limited affinity for halides in the gas- H tBu H O H O t t H N O Bu phase and therefore no noticeable binding was observed in Bu n t N N O + N Bu4X Bu H N O N tBu solution, where intraligand H-bonding interaction are believe N N tBu N to be predominant.25, 27 Protonation of the central nitrogen disturbs this well-organized network, and in an appropriate Scheme 3. Halide binding equilibria for L(HX), and free L. solvent (CH2Cl2, CHCl3), lead to discrete tripodal anion receptor as crystallized for L(HBr)2(CHCl3) (Figure 4C). However To further confirm the stoichiometry of the binding model, we protonation of L to [L(H)]+ creates a situation where the performed 1H DOSY NMR studies which demonstrated a approximate C3-symmetric tripodal conformer is competed by n similar diffusion coefficient for L(HBr) and [L(HBr2)]N Bu4, this a “Janus head” form as highlighted by the computational suggests that, at the NMR timescale, no larger oligomer is results. This effect is illustrated by multiple lines of evidence, formed, which allows us to further rule out a potential 2:1 such as the propensity of L(HX) to bind a second halide in − binding model. Interestingly, titration experiments between L solution, the high ESI yield of [L(HX2)] , the CID results or the n and [N Bu4]X did not reveal any noticeable binding (Scheme 3). exothermic second halide binding as determined by DFT. From

Although there is about a 4-fold difference between KCl and KI there, aggregation of multiple units can start as corroborated − + (which is small compared to other systems),16, 62 this only by the observation of [(L)2(H2X3)] and [(L)2(H2X)] by ESI.

corresponds to an energy difference, ΔΔG = RTln[KCl/KI], of Moreover, the crystallization of L(HI)½EtOH is a remarkable ~0.8 kcal mol−1, which is small relative to differences in gas- example of an arrested aggregation step with both the phase energetics discussed above. These binding equilibria “tripodal” and “Janus head” forms of [L(H)]+ present (Figure were also evaluated by DFT methods that are generally 4B). The processes can then be repeated indefinitely to yield consistent with the experimental trend (Table S8). The affinity the 1D-H-bonded polymers crystallized for L(HCl) and L(HBr). of neutral L for halide anions was also assessed by DFT Overall, the complementary findings between multiple calculations, demonstrating a similar trend as that of L(HX) but techniques allow us to identify and monitor in detail the key with much weaker association energies (Table S10). Overall, molecular steps underlying the specific

Please do not adjust margins PhysicalPlease Chemistry do not adjust Chemical margins Physics Page 8 of 11

ARTICLE Journal Name precipitation/crystallization of [L(H)]+ (Scheme 4) but may be Affinity of protonated tripodal hydroxylamine ligand [L(H)]+ for generalized to a large array of molecules. halides X− was suggested in ESI by abundant gas-phase − complexes [L(HX2)] . Relative yields from solutions containing HSol − SolH X H O more than one halide indicated a preferential affinity for I N − − − H over Br over Cl . CID of mixed halide complexes [L(HXaXb)] "Janus head" SolH HSol N form H X H O N also revealed preferential retention of the heavier halide in O H N O N H H O H − N O O O O O H [L(X )] via elimination of HX . Computed DFT energies are in N N HX + [L(HX)] H + n [L(HX)] X b a H X N X N H HSol N N H H – [L(HX)] H H O accord with the observed gas-phase speciation and CID N H N H Sol = CHCl , O O O N 2 N O n+2 H3TriNOx (L) CCl3, OEt "Tripodal" fragmentation pathways. Energetics reveal that preferential [L(HX)] N N N form H "Janus head" + form retention of the heavier halide Xb by [L(H)] does not reflect N Scheme 4. Schematic oligomerization process for L(HX). H-bonding interactions intrinsically higher affinity but rather higher stability of the are depicted in pink. lighter HXa product. Halide affinity of L(HX) was confirmed by binding equilibria From Solution to Gas by ESI constants in solution. The solution results indicated the highest Given that ESI resulted in gas-phase complexes that bear affinity for Cl−, a lower affinity for Br−, and the lowest affinity compositional correspondence to condensed phase species, for I−, which is the opposite of what is observed in gas phase we briefly address the nature of ESI vis-à-vis solution but is in accord with gas-phase affinities obtained after speciation. ESI is often considered a “soft” ionization method, accounting for stabilities of gas-phase HCl, HBr and HI. The 1D largely due to its ability to transfer intact covalently bonded polymeric structures of solid L(HX) exhibit a remarkable macromolecules from solution to gas, as pioneered by Fenn.63 correspondence to the compositions of gas-phase complexes However, transferring a charged species from solution to gas produced by ESI. The solid structures also bear a close necessarily involves drastic changes in the transition from resemblance to computed gas-phase structures. The results bulk- to micro- to nano- to molecular-“solution” environments, suggest hydroxylamines and related substrates as potentially with concomitant opportunities for changes in chemistry, promising for anion reception and recognition. including in speciation.64 Potential pitfalls in inferring solution speciation from ESI-MS have been emphasized in recent years.65-68 There are examples of judicious and effective Conflicts of interest application of ESI-MS to assess condensed phase structures There are no conflicts to declare. and reactivity of non-covalently bound systems such as supramolecular containers.69 Highly charged solution metal ions, M4+, were transferred from solution to gas, but only Acknowledgements when stabilized against hydrolysis and charge-reduction by Research at UPenn, Berkeley Lab and Los Alamos was multidentate coordinating ligands.70-72 supported as part of the Center for Actinide Science and Among the dynamic effects during ESI are drastic changes in Technology (CAST), an Energy Frontier Research Center (EFRC) ion concentrations, including pH.73 Because solution species funded by the U.S. Department of Energy (DOE), Office of that are precursors of the solid 1D-H-bonded coordination Science, Basic Energy Sciences (BES), under Award Number DE- polymer are acid adducts of L, L(HX), it is expected that pH SC0016568. Theoretical research was performed using EMSL changes during ESI, as well halide concentration changes, (grid.436923.9), a DOE Office of Science User Facility could affect compositions of gas-phase species. For example, a sponsored by the Office of Biological and Environmental decrease in pH should generally result in an increase in Research. We acknowledge Prof. Polly L. Arnold (University of concentration of the associated weaker acid HI, which could Edinburgh) for insightful discussions during the development increase the concentration of neutral L(HI) and anionic − of this work. [L(HI2)] . Results such as in Figure 1 may thus reflect aspects of solution speciation, but cannot be taken to directly reveal it. It cannot be concluded from ESI that L is selective for heavier Notes and references halides in solution, but rather that such selectivity is exhibited 1. Kirk, K. L., Biochemistry of the Elemental Halogens and Inorganic in ESI. Halides. Springer US: 2012. CID does demonstrate propensity for particular fragmentation 2. Olander, D. R., The theory of uranium enrichment by the gas pathways. However, preferred elimination of a particular centrifuge. Progress in Nuclear Energy 1981, 8 (1), 1-33. halide needs to be interpreted in the proper context, such as 3. Metcalfe, B. L.; Donald, I. W.; Fong, S. K.; Gerrard, L. A.; by comparing energetics for reactions (4) and (6). The overall Strachan, D. M.; Scheele, R. D., Ageing of a phosphate ceramic used assessment here is that CID does not necessarily indicate to immobilize chloride contaminated actinide waste. J. Nucl. Mater. inherently preferential binding of heavier halides, but rather 2009, 385 (2), 485-488. higher stability of lighter hydrogen halides. 4. Fong, S. K.; Donald, I. W.; Metcalfe, B. L., Development of a glass-encapsulated calcium phosphate wasteform for the immobilization of actinide and halide containing radioactive wastes Conclusions

Please do not adjust margins Page 9 of 11 PhysicalPlease Chemistry do not adjust Chemical margins Physics

Journal Name ARTICLE

from the pyrochemical reprocessing of plutonium metal. J. Alloys 24. Brown, A.; Beer, P. D., Halogen bonding anion recognition. Compd. 2007, 444-445, 424-428. Chem. Commun. 2016, 52 (56), 8645-8658. 5. Riley, B. J.; Vienna, J. D.; Strachan, D. M.; McCloy, J. S.; Jerden, J. 25. Bogart, J. A.; Lippincott, C. A.; Carroll, P. J.; Schelter, E. J., An L., Materials and processes for the effective capture and Operationally Simple Method for Separating the Rare-Earth immobilization of radioiodine: A review. J. Nucl. Mater. 2016, 470, Elements Neodymium and Dysprosium. Angewandte Chemie 307-326. International Edition 2015, 54 (28), 8222-8225. 6. Ojovan, M. I.; Lee, W. E., An Introduction to Nuclear Waste 26. Fang, H.; Cole, B. E.; Qiao, Y.; Bogart, J. A.; Cheisson, T.; Manor, Immobilisation. Elsevier Science: 2013. B. C.; Carroll, P. J.; Schelter, E. J., Electro-kinetic Separation of Rare 7. Hrma, P. R., Retention of Halogen in Waste Glass. U.S. Earth Elements Using a Redox-Active Ligand. Angew. Chem. Int. Ed. Department of Energy: Pacific Northwest National Laboratory, 2017, 56 (43), 13450-13454. 2010. 27. Cole, B. E.; Falcones, I. B.; Cheisson, T.; Manor, B.; Carroll, P.; 8. Gin, S.; Abdelouas, A.; Criscenti, L. J.; Ebert, W. L.; Ferrand, K.; Schelter, E. J., A Molecular Basis to Rare Earth Separations for Geisler, T.; Harrison, M. T.; Inagaki, Y.; Mitsui, S.; Mueller, K. T.; Recycling: Tuning TriNOx Ligand Properties for Improved Marra, J. C.; Pantano, C. G.; Pierce, E. M.; Ryan, J. V.; Schofield, J. Performance. Chem. Commun. 2018, 54, 10276–10279. M.; Steefel, C. I.; Vienna, J. D., An international initiative on long- 28. Cheisson, T.; Solola, L. A.; Gau, M. R.; Carroll, P. J.; Schelter, E. J., term behavior of high-level nuclear waste glass. Mater. Today 2013, Silyl Transfer Pathway to a Ce(IV) Imido Complex. Organometallics 16 (6), 243-248. 2018, 37 (23), 4332-4335. 9. Vienna, J. D., Nuclear Waste Vitrification in the United States: 29. McSkimming, A.; Su, J.; Cheisson, T.; Gau, M. R.; Carroll, P. J.; Recent Developments and Future Options. International Journal of Batista, E. R.; Yang, P.; Schelter, E. J., Coordination Chemistry of a Applied Glass Science 2010, 1 (3), 309-321. Strongly-Donating Hydroxylamine with Early Actinides: An 10. Peterson, R. A.; Buck, E. C.; Chun, J.; Daniel, R. C.; Herting, D. L.; Investigation of Redox Properties and Electronic Structure. Inorg. Ilton, E. S.; Lumetta, G. J.; Clark, S. B., Review of the Scientific Chem. 2018, 57 (8), 4387-4394. Understanding of Radioactive Waste at the U.S. DOE Hanford Site. 30. Cheisson, T.; Cole, B. E.; Manor, B. C.; Carroll, P. J.; Schelter, E. Environ. Sci. Technol. 2018, 52 (2), 381-396. J., Phosphoryl-Ligand Adducts of Rare Earth-TriNOx Complexes: 11. Noyes, R., Nuclear Waste Cleanup Technologies and Systematic Studies and Implications for Separations Chemistry. ACS Opportunities. Elsevier Science: 2013. Sustainable Chem. Eng. 2019, 7 (5), 4993-5001. 12. Billings, A. L.; Fox, K. M., Retention of Sulfate in Savannah River 31. Cole, B. E.; Cheisson, T.; Higgins, R. F.; Nakamaru-Ogiso, E.; Site High-Level Radioactive Waste Glass. International Journal of Manor, B. C.; Carroll, P. J.; Schelter, E. J., Redox-Driven Chelation Applied Glass Science 2010, 1 (4), 388-400. and Kinetic Separation of Select Rare Earths Using a Tripodal 13. Park, C. H.; Simmons, H. E., Macrobicyclic amines. III. Nitroxide Proligand. Inorg. Chem. 2019. Encapsulation of halide ions by in,in-1,(k + 2)- 32. Shannon, R., Revised effective ionic radii and systematic studies diazabicyclo[k.l.m.]alkane ammonium ions. J. Am. Chem. Soc. 1968, of interatomic distances in halides and chalcogenides. Acta 90 (9), 2431-2432. Crystallogr. Sect. A 1976, 32 (5), 751-767. 14. Graf, E.; Lehn, J. M., Anion cryptates: highly stable and selective 33. Rahm, M.; Zeng, T.; Hoffmann, R., Electronegativity Seen as the macrotricyclic anion inclusion complexes. J. Am. Chem. Soc. 1976, Ground-State Average Valence Electron Binding Energy. J. Am. 98 (20), 6403-6405. Chem. Soc. 2019, 141 (1), 342-351. 15. Sessler, J. L.; Gale, P.; Cho, W. S., Anion Receptor Chemistry. 34. Eller, K.; Schwarz, H., Organometallic Chemistry in the Gas- Royal Society of Chemistry: 2007. Phase. Chem. Rev. 1991, 91 (6), 1121-1177. 16. He, Q.; Tu, P.; Sessler, J. L., Supramolecular Chemistry of Anionic 35. Schalley, C. A.; Hornung, G.; Schroder, D.; Schwarz, H., Mass Dimers, Trimers, Tetramers, and Clusters. Chem 2018, 4 (1), 46-93. spectrometric approaches to the reactivity of transient neutrals. 17. Ayme, J.-F.; Beves, J. E.; Campbell, C. J.; Gil-Ramírez, G.; Leigh, Chem. Soc. Rev. 1998, 27 (2), 91-104. D. A.; Stephens, A. J., Strong and Selective Anion Binding within the 36. Meot-Ner, M., Update 1 of: Strong Ionic Hydrogen Bonds. Central Cavity of Molecular Knots and Links. J. Am. Chem. Soc. 2015, Chem. Rev. 2012, 112 (10), PR22-PR103. 137 (31), 9812-9815. 37. O'Hair, R. A. J., The 3D quadrupole ion trap mass spectrometer 18. Liu, Y.; Sengupta, A.; Raghavachari, K.; Flood, A. H., Anion as a complete chemical laboratory for fundamental gas-phase Binding in Solution: Beyond the Electrostatic Regime. Chem 2017, 3 studies of metal mediated chemistry. Chem. Commun. 2006, (14), (3), 411-427. 1469-1481. 19. Busschaert, N.; Caltagirone, C.; Van Rossom, W.; Gale, P. A., 38. Legon, A. C., The Nature of Ammonium and Methylammonium Applications of Supramolecular Anion Recognition. Chem. Rev. Halides in the Vapor-Phase - Hydrogen-Bonding Versus Proton- 2015, 115 (15), 8038-8155. Transfer. Chem. Soc. Rev. 1993, 22 (3), 153-163. 20. Langton, M. J.; Serpell, C. J.; Beer, P. D., Anion Recognition in 39. Walker, B. W.; Sunderlin, L. S., The thermochemistry of formic Water: Recent Advances from a Supramolecular and acid halide anion clusters. Int. J. Mass spectrom. 1999, 184 (2-3), Macromolecular Perspective. Angew. Chem. Int. Ed. 2016, 55 (6), 183-189. 1974-1987. 40. Bogdanov, B.; McMahon, T. B., An ab initio and density 21. Ahmed, L.; Rhaman, M. M.; Mendy, J. S.; Wang, J.; Fronczek, F. functional theory investigation of the structures and energetics of R.; Powell, D. R.; Leszczynski, J.; Hossain, M. A., Experimental and halide ion-alcohol complexes in the gas phase. J. Phys. Chem. A Theoretical Studies on Halide Binding with a p-Xylyl-Based 2000, 104 (33), 7871-7880. Azamacrocycle. J. Phys. Chem. A 2015, 119 (2), 383-394. 41. Hiraoka, K.; Katsuragawa, J.; Sugiyama, T.; Kojima, T.; Yamabe, 22. Gale, Philip A.; Howe, Ethan N. W.; Wu, X., Anion Receptor S., Hydrogen bonds in gas-phase clusters between halide ions and Chemistry. Chem 2016, 1 (3), 351-422. olefins. J. Am. Soc. Mass. Spectrom. 2001, 12 (2), 144-149. 23. Liu, Y.; Zhao, W.; Chen, C.-H.; Flood, A. H., Chloride capture 42. Gillis, E. A. L.; Demireva, M.; Sarwar, M. G.; Chudzinski, M. G.; using a C–H hydrogen bonding cage. Science 2019, eaaw5145. Taylor, M. S.; Williams, E. R.; Fridgen, T. D., Structure and energetics

Please do not adjust margins PhysicalPlease Chemistry do not adjust Chemical margins Physics Page 10 of 11

ARTICLE Journal Name of gas phase halogen-bonding in mono-, bi-, and tri-dentate anion 59. Pramanik, A.; Powell, D. R.; Wong, B. M.; Hossain, M. A., receptors as studied by BIRD. Phys. Chem. Chem. Phys. 2013, 15 Spectroscopic, Structural, and Theoretical Studies of Halide (20), 7638-7647. Complexes with a Urea-Based Tripodal Receptor. Inorg. Chem. 43. Chiavarino, B.; Crestoni, M. E.; Maitre, P.; Fornarini, S., Halide 2012, 51 (7), 4274-4284. adducts of 1,3,5-trinitrobenzene: Vibrational signatures and role of 60. Ulatowski, F.; Dąbrowa, K.; Bałakier, T.; Jurczak, J., Recognizing anion-pi interactions. Int. J. Mass spectrom. 2013, 354, 62-69. the Limited Applicability of Job Plots in Studying Host–Guest 44. O'Brien, J. T.; Prell, J. S.; Berden, G.; Oomens, J.; Williams, E. R., Interactions in Supramolecular Chemistry. J. Org. Chem. 2016, 81 Effects of anions on the zwitterion stability of Glu, His and Arg (5), 1746-1756. investigated by IRMPD spectroscopy and theory. Int. J. Mass 61. Thordarson, P., Determining association constants from spectrom. 2010, 297 (1-3), 116-123. titration experiments in supramolecular chemistry. Chem. Soc. Rev. 45. Corinti, D.; Gregori, B.; Guidoni, L.; Scuderi, D.; McMahon, T. B.; 2011, 40 (3), 1305-1323. Chiavarino, B.; Fornarini, S.; Crestoni, M. E., Complexation of halide 62. de Namor, A. F. D.; Shehab, M., Selective recognition of halide ions to tyrosine: role of non-covalent interactions evidenced by anions by calix[4]pyrrole: A detailed thermodynamic study. J. Phys. IRMPD spectroscopy. Phys. Chem. Chem. Phys. 2018, 20 (6), 4429- Chem. B 2003, 107 (26), 6462-6468. 4441. 63. Whitehouse, C. M.; Dreyer, R. N.; Yamashita, M.; Fenn, J. B., 46. Cheisson, T.; Schelter, E. J., Rare Earths Elements: Mendeleev’s Electrospray Interface for Liquid Chromatographs and Mass Bane, Modern Marvels. Science 2019, 363, 489–493. Spectrometers. Anal. Chem. 1985, 57 (3), 675-679. 47. Rutkowski, P. X.; Michelini, M. C.; Gibson, J. K., Gas-phase 64. Kebarle, P.; Verkerk, U. H., Electrospray: From Ions in Solution lanthanide chloride clusters: relationships among ESI abundances to Ions in the Gas Phase, What We Know Now. Mass Spectrom. Rev. and DFT structures and energetics. Phys. Chem. Chem. Phys. 2012, 2009, 28 (6), 898-917. 14 (6), 1965-1977. 65. Di Marco, V. B.; Bombi, G. G.; Zambon, S.; Traidi, P., Metal- 48. Dau, P. D.; Gibson, J. K., Halide Abstraction from Halogenated ligand solution equilibria studied by electrospray ionization mass Acetate Ligands by Actinyls: A Competition between Bond Breaking spectrometry: effect of instrumental parameters. J. Mass Spectrom. and Bond Making. J. Phys. Chem. A 2015, 119 (13), 3218-3224. 2009, 44 (1), 120-127. 49. Lias, S. G.; Bartmess, J. E.; Liebman, J. F.; Holmes, J. L.; Levin, R. 66. Di Marco, V. B.; Raveane, L.; Dean, A.; Traldi, P., Perturbations D.; Mallard, W. G., Gas-Phase Ion and Neutral Thermochemistry. J. produced by electrospray ionization mass spectrometry in the Phys. Chem. Ref. Data 1988, 17 (1), 1-861. speciation of aluminium(III)/1,6-dimethy1-4-hydroxy-3- 50. Blanksby, S. J.; Ellison, G. B., Bond Dissociation Energies of pyridinecarboxylate aqueous solutions. Rapid Commun. Mass Organic Molecules. Acc. Chem. Res. 2003, 36 (4), 255-263. Spectrom. 2010, 24 (7), 868-874. 51. Cooks, R. G.; Wong, P. S. H., Kinetic method of making 67. McDonald, L. W.; Campbell, J. A.; Clark, S. B., Failure of ESI thermochemical determinations: Advances and applications. Spectra to Represent Metal-Complex Solution Composition: A Study Accounts of Chemical Research 1998, 31 (7), 379-386. of Lanthanide-Carboxylate Complexes. Anal. Chem. 2014, 86 (2), 52. Ellis, R. J.; Chartres, J.; Henderson, D. K.; Cabot, R.; Richardson, 1023-1029. P. R.; White, F. J.; Schröder, M.; Turkington, J. R.; Tasker, P. A.; Sole, 68. Swadzba-Kwasny, M., Comment on "Lewis acidic ionic liquids of K. C., Design and Function of Pre-organised Outer-Sphere crown ether complex cations: preparation and applications in Amidopyridyl Extractants for Zinc(II) and Cobalt(II) organic reactions" by Y. Liang, J. Wang, C. Cheng and H. Jing, RSC Chlorometallates: The Role of C–H Hydrogen Bonds. Chem. Eur. J. Adv., 2016, 6, 93546. RSC Adv. 2017, 7 (82), 51907-51909. 2012, 18 (25), 7715-7728. 69. Qi, Z. H.; Heinrich, T.; Moorthy, S.; Schalley, C. A., Gas-phase 53. Turkington, J. R.; Bailey, P. J.; Love, J. B.; Wilson, A. M.; Tasker, chemistry of molecular containers. Chem. Soc. Rev. 2015, 44 (2), P. A., Exploiting outer-sphere interactions to enhance metal 515-531. recovery by solvent extraction. Chem. Commun. 2013, 49 (19), 70. Gong, Y.; Hu, H. S.; Tian, G. X.; Rao, L. F.; Li, J.; Gibson, J. K., A 1891-1899. Tetrapositive Metal Ion in the Gas Phase: Thorium(IV) Coordinated 54. Carson, I.; MacRuary, K. J.; Doidge, E. D.; Ellis, R. J.; Grant, R. A.; by Neutral Tridentate Ligands. Angew Chem Int Edit 2013, 52 (27), Gordon, R. J.; Love, J. B.; Morrison, C. A.; Nichol, G. S.; Tasker, P. A.; 6885-6888. Wilson, A. M., Anion Receptor Design: Exploiting Outer-Sphere 71. Gong, Y.; Tian, G. X.; Rao, L. F.; Gibson, J. K., Tetrapositive Coordination Chemistry To Obtain High Selectivity for Plutonium, Neptunium, Uranium, and Thorium Coordination Chloridometalates over Chloride. Inorg. Chem. 2015, 54 (17), 8685- Complexes: Chemistry Revealed by Electron Transfer and Collision 8692. Induced Dissociation. J. Phys. Chem. A 2014, 118 (15), 2749-2755. 55. Al-Shnani, F.; Guisado-Barrios, G.; Sainz, D.; Peris, E., Tris- 72. Rutkowski, P. X.; Michelini, M. D.; Gibson, J. K., Proton Transfer triazolium Salts as Anion Receptors and as Precursors for the in Th(IV) Hydrate Clusters: A Link to Hydrolysis of Th(OH)(2)(2+) to Preparation of Cylinder-like Coordination Cages. Organometallics Th(OH)(3)(+) in Aqueous Solution. J. Phys. Chem. A 2013, 117 (2), 2019, 38 (3), 697-701. 451-459. 56. Sengupta, A.; Liu, Y.; Flood, A. H.; Raghavachari, K., Anion- 73. Konermann, L., Addressing a Common Misconception: Binding Macrocycles Operate Beyond the Electrostatic Regime: Ammonium Acetate as Neutral pH "Buffer" for Native Electrospray Interaction Distances Matter. Chem. Eur. J. 2018, 24 (54), 14409- Mass Spectrometry. J. Am. Soc. Mass. Spectrom. 2017, 28 (9), 1827- 14417. 1835. 57. Nepal, B.; Scheiner, S., Competitive Halide Binding by Halogen Versus Hydrogen Bonding: Bis-triazole Pyridinium. Chem. Eur. J. 2015, 21 (38), 13330-13335. 58. Lee, S. J. R.; Mullinax, J. W.; Schaefer, H. F., Intermolecular interactions and proton transfer in the hydrogen halide–superoxide anion complexes. Phys. Chem. Chem. Phys. 2016, 18 (8), 6201-6208.

Please do not adjust margins Page 11 of 11 PhysicalPlease Chemistry do not adjust Chemical margins Physics

Journal Name ARTICLE

The binding of halide anions with a tripodal hydroxylamine ligand studied in gas (mass spectrometry and DFT methods) and condensed phases revealed notable agreement.

Please do not adjust margins